Semantic Textual Similarity On Sts Benchmark

评估指标

Spearman Correlation

评测结果

各个模型在此基准测试上的表现结果

		Paper Title	Repository
Mnet-Sim	0.931	MNet-Sim: A Multi-layered Semantic Similarity Network to Evaluate Sentence Similarity	-
MT-DNN-SMART	0.925	SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization
StructBERTRoBERTa ensemble	0.924	StructBERT: Incorporating Language Structures into Pre-training for Deep Language Understanding	-
T5-11B	0.921	Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
RealFormer	0.8988	RealFormer: Transformer Likes Residual Attention
T5-3B	0.898	Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
AnglE-LLaMA-13B	0.8969	AnglE-optimized Text Embeddings
ASA + RoBERTa	0.892	Adversarial Self-Attention for Language Understanding
PromptEOL+CSE+LLaMA-30B	0.8914	Scaling Sentence Embeddings with Large Language Models
AnglE-LLaMA-7B	0.8897	AnglE-optimized Text Embeddings
AnglE-LLaMA-7B-v2	0.8897	AnglE-optimized Text Embeddings
T5-Large 770M	0.886	Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
PromptEOL+CSE+OPT-13B	0.8856	Scaling Sentence Embeddings with Large Language Models
PromptEOL+CSE+OPT-2.7B	0.8833	Scaling Sentence Embeddings with Large Language Models
PromCSE-RoBERTa-large (0.355B)	0.8787	Improved Universal Sentence Embeddings with Prompt-based Contrastive Learning and Energy-based Learning
BigBird	.878	Big Bird: Transformers for Longer Sequences
Trans-Encoder-RoBERTa-large-cross (unsup.)	0.867	Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations
SimCSE-RoBERTalarge	0.867	SimCSE: Simple Contrastive Learning of Sentence Embeddings
Trans-Encoder-RoBERTa-large-bi (unsup.)	0.8655	Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations
ASA + BERT-base	0.865	Adversarial Self-Attention for Language Understanding

0 of 66 row(s) selected.